Computational approaches to understand transcription regulation in development

Abstract

Gene regulatory networks (GRNs) serve as useful abstractions to understand transcriptional dynamics in developmental systems. Computational prediction of GRNs has been successfully applied to genome-wide gene expression measurements with the advent of microarrays and RNA-sequencing. However, these inferred networks are inaccurate and mostly based on correlative rather than causative interactions. In this review, we highlight three approaches that significantly impact GRN inference: (1) moving from one genome-wide functional modality, gene expression, to multi-omics, (2) single cell sequencing, to measure cell type-specific signals and predict context-specific GRNs, and (3) neural networks as flexible models. Together, these experimental and computational developments have the potential to significantly impact the quality of inferred GRNs. Ultimately, accurately modeling the regulatory interactions between transcription factors and their target genes will be essential to understand the role of transcription factors in driving developmental gene expression programs and to derive testable hypotheses for validation.

Single-cell DNA Methylome and 3D Multi-omic Atlas of the Adult Mouse Brain

“…We first performed quantile normalization along subclasses using the Python package qnorm (v0.8.0) 95 to normalize the mC fractions and compartment scores. We then calculated the PCC between the compartment scores of nonoverlapping chromosome 100Kb bins with the corresponding bin's mCH or mCG fractions across cell subclasses.…”

“…We first performed quantile normalization along subclasses using the Python package qnorm (v0.8.0) 95 to normalize the transcript body mC fractions and chromosome 25Kb bin boundary probabilities. We then calculated the PCC between the differential boundary probabilities of 25Kb bins with the transcript body mCH and mCG fractions.…”

“…To investigate the relationship between gene status and the surrounding chromatin conformation diversity, we first performed quantile normalization along subclasses using the Python package qnorm (v0.8.0) 95 to normalize the transcript body mCH fractions and contact strengths of highly variable interactions. We then calculated PCC between the transcript body mCH fraction and the highly variable interactions if any anchor of the interactions had overlapped with the gene body.…”

“…The basis of each pairwise edge was the correlation between the methylation fractions of the two genome elements across cell subclasses. We performed quantile normalization along subclasses using the Python package qnorm (v0.8.0) 95 to normalize the two matrices involved in calculating the correlation. Gene body mCH fraction was used as a proxy for TF and target gene activity, and mCG fractions were used to represent DMR status.…”

Cracking the Code of Neuronal Cell Fate

“…Thus, cracking the code of neuronal fate may elicit novel pharmacological strategies, no longer oriented to the cellular hardware but, rather, the nuclear transcriptional regulatory mechanisms. Reconstruction of this cellular program by “reverse engineering” of gene regulatory networks (GRNs) poses great opportunities in systems biology [ 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 , 96 , 97 ] and allows to build accurate models of physiological and pathological processes, including those implicated in neuronal fate and development [ 98 , 99 , 100 , 101 , 102 , 103 ]. The impact of using these gene regulatory models to understand human diseases and find new treatments is profound, since they may allow to identify disease driver genes and promising biomarkers and therapeutic targets more efficiently and accurately [ 99 , 103 , 104 , 105 , 106 ].…”